Renaming `affine_hull` for polyhedra

[Jonathan Kliem]

Currently, the method `affine_hull` returns a polyhedron projected into its affine hull.

However, the affine hull of a polyhedron is the smallest affine set containing it.

Therefore, I propose to rename the method:

affine_hull -> in_affine_hull

in #29326.

Do you have thoughts about it? Does anyone have a better name? Opinions are welcome.

Thanks,

Jonathan

[Dima Pasechnik]

affine_hull_projection

Thanks,

Jonathan

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/36c4dbc3-c428-4617-8407-b1b8e3e77c59%40googlegroups.com.

[Vincent Delecroix]

Le 13/03/2020 à 17:49, Dima Pasechnik a écrit :
> On Fri, 13 Mar 2020, 16:46 'Jonathan Kliem' via sage-devel, <
> sage-...@googlegroups.com> wrote:
>
>> Currently, the method `affine_hull` returns a polyhedron projected into
>> its affine hull.
>> However, the affine hull of a polyhedron is the smallest affine set
>> containing it.
>>
>> Therefore, I propose to rename the method:
>>
>> affine_hull -> in_affine_hull
>>

>> in #29326. <https://trac.sagemath.org/ticket/29326>

>>
>> Do you have thoughts about it? Does anyone have a better name? Opinions
>> are welcome.
>>
>
> affine_hull_projection

This is indeed much better.

[Jonathan Kliem]

I agree.

[Matthias Koeppe]

+1 

[jplab]

in #29326.

Do you have thoughts about it? Does anyone have a better name? Opinions are welcome.

affine_hull_projection

+1 for affine_hull_projection

At one moment, I thought something called "make_full_dimensional" or would be a very straightforward name. One could think of eventually add such an alias, but I would be already happy with "affine_hull_projection".

[Vincent Delecroix]

Le 15/03/2020 à 16:21, jplab a écrit :
>
>>
>>
>>> in #29326. <https://trac.sagemath.org/ticket/29326>

make_full_dimensional is ambiguous. For any complementary subspace L
to the affine hull polytope x L is full dimensional in the ambient
space. That would also be a perfectly valid construction for
make_full_dimensional.

-1 for the alias